Combined average of Moat (AI Resilience), Growth, and Valuation scores.
Moat Score
One of the two largest US mechanical contractors, now with a fast-growing electrical arm and a modular prefabrication footprint heading past 4 million square feet — a moat built on scarce craft labour, repeat relationships with hyperscale and advanced-manufacturing builders, and execution that earns materially higher margins than its closest rival.
Comfort Systems' edge is execution on the hardest building systems in the country — mechanical and electrical scopes for data centres, chip fabs and pharma plants, increasingly built off-site in its own modular plants. It is a real, margin-proven advantage, but one that rests on people and process rather than on a structural lock:
- Margin Leadership Over the Closest Rival: In Q2 2026 Comfort Systems earned ~17.1% operating margin on $3.27B of revenue, against EMCOR's 10.6% on $5.15B. Gross margin rose to 25.9% from 23.5%, with Mechanical at 25.6% and Electrical at 26.4%. The gap reflects mix toward technology work, decentralised operating companies that price their own jobs, and fewer construction-manager or prime-contractor roles — EMCOR's mechanical margin fell 110 bp on exactly that shift.
- Modular Prefabrication: The modular business booked a record ~$510M of new orders in Q2, and management is expanding capacity to more than 4 million square feet by the end of 2026 and about 5 million by late summer 2027. Building skids, racks and electrical rooms in a factory moves craft hours off congested job sites, which is the scarce input in a data-centre build.
- Embedded With Technology Builders: Technology customers, including data centres, were about 58% of revenue in the first half of 2026, up from about 40%. Repeat campus work with the same builders turns into backlog visibility — $14.06B, roughly 1.1× this year's expected revenue — but also concentrates the book in a handful of hyperscaler programmes.
Moat Verdict
Comfort Systems is a net AI beneficiary: data-centre and chip-fab construction drove technology work to about 58% of revenue and backlog to a record $14.06B, and nothing in its physical, craft-labour business is exposed to AI substitution. Its advantages — scarce trades, modular prefabrication, repeat relationships and a margin lead over EMCOR — are real but all intact rather than strong, so the franchise is durable through this cycle while the main risk is that the cycle itself digests.
65.0 resilient · 65.0 vulnerable · 80/20 = 65.0 · = 65
Open a moat to read its note.
Construction services contractor; there is no end-user interface to learn.
Mechanical, electrical and modular build know-how is physical execution, not software business logic (M7, the Cameco precedent). Quanta rates this pillar weakened for internal project tooling; under M7 that looks like the one to revisit, not this.
No public-data moat.
Skilled pipefitters, sheet-metal workers and electricians are the binding constraint on data-centre builds, and the decentralised operating companies retain local foremen and estimators well. Rivals such as EMCOR hire from the same craft pool and no captive training pipeline or headcount figure is on file, so it rates intact on M8 — unlike Quanta, which is strong on its captive lineman college.
Mechanical, electrical and factory-built modular scopes let it bid integrated building-systems packages, and Electrical (+81% in Q2) is being cross-sold into mechanical customers. No attach or retention figure is on file, so it rates intact on M9, as Quanta's integrated EPC scope does.
Estimating and job-cost history is internal tooling, not a data asset with a moat. Quanta rates the same class of bid data weakened; under M6 a pillar that never applied is na, so this follows the rule rather than that rating.
State contractor licences, safety records and customer prequalification bar small entrants, but EMCOR and other national contractors clear the same hurdles, so it rates intact on M4, as Quanta's utility prequalification does.
Services contractor with no network effect.
Repeat campus work for hyperscale and advanced-manufacturing builders, plus a service and retrofit business on installed systems, makes it a first-call contractor. Jobs are still bid project by project and EMCOR holds the same relationships, so it rates intact, as Quanta's MSAs do.
Not a system of record.
Modular capacity heading past 4 million square feet and Q2 operating margin of ~17.1% against EMCOR's 10.6% show a real execution and mix advantage, but EMCOR is the larger contractor ($5.15B quarterly revenue vs $3.27B), so the lead is not a unit-cost lead from scale. Intact, as for Quanta.
B2B contractor chosen on capacity, schedule and execution, which the other pillars rate; the name carries reputation, not a price premium it could hold on brand alone (M10).
Combined average of Moat (AI Resilience), Growth, and Valuation scores.
Moat Score
One of the two largest US mechanical contractors, now with a fast-growing electrical arm and a modular prefabrication footprint heading past 4 million square feet — a moat built on scarce craft labour, repeat relationships with hyperscale and advanced-manufacturing builders, and execution that earns materially higher margins than its closest rival.
Growth Score
Q2 2026 revenue was $3.27B, +50% YoY (same-store +44%), with Mechanical +40% to $2.30B and Electrical +81% to $969M ($132.6M of that increase acquired, largely Hunt Electric, closed May 1). Q2 diluted EPS was $12.53 vs $6.53; operating income $558M vs $299.9M. Backlog reached a record $14.06B (+73% YoY; $5.6B of the increase same-store), and management raised its 2026 same-store revenue growth outlook to the mid-to-high 30% range from mid-to-high 20%, while flagging tougher fourth-quarter comparisons. FY2025 revenue was $9.10B (+29%) with EPS of $28.88.
Valuation Score
At ~$1,715 (October 9, 2026 close) — about 17% below the $2,073.99 52-week high and more than double the ~$770 low — FIX trades at ~42× trailing EPS (~$40.6), ~35× a 2026 EPS of roughly $49–50 (first-half actual $23.03 plus the Street's second half) and ~28× the ~$61 2027 consensus. On a $1,250 / $1,900 / $2,600 ladder the price sits about 72% of the way from bear to base, giving a valuation score of 72: below a base set at a PWR-like ~31× 2027 EPS and ~9% under the Street mean, with Q3 results due October 22.
The Craft-Labour and Modular-Execution Moat
Comfort Systems' edge is execution on the hardest building systems in the country — mechanical and electrical scopes for data centres, chip fabs and pharma plants, increasingly built off-site in its own modular plants. It is a real, margin-proven advantage, but one that rests on people and process rather than on a structural lock:
- Margin Leadership Over the Closest Rival: In Q2 2026 Comfort Systems earned ~17.1% operating margin on $3.27B of revenue, against EMCOR's 10.6% on $5.15B. Gross margin rose to 25.9% from 23.5%, with Mechanical at 25.6% and Electrical at 26.4%. The gap reflects mix toward technology work, decentralised operating companies that price their own jobs, and fewer construction-manager or prime-contractor roles — EMCOR's mechanical margin fell 110 bp on exactly that shift.
- Modular Prefabrication: The modular business booked a record ~$510M of new orders in Q2, and management is expanding capacity to more than 4 million square feet by the end of 2026 and about 5 million by late summer 2027. Building skids, racks and electrical rooms in a factory moves craft hours off congested job sites, which is the scarce input in a data-centre build.
- Embedded With Technology Builders: Technology customers, including data centres, were about 58% of revenue in the first half of 2026, up from about 40%. Repeat campus work with the same builders turns into backlog visibility — $14.06B, roughly 1.1× this year's expected revenue — but also concentrates the book in a handful of hyperscaler programmes.
Moat Verdict
Comfort Systems is a net AI beneficiary: data-centre and chip-fab construction drove technology work to about 58% of revenue and backlog to a record $14.06B, and nothing in its physical, craft-labour business is exposed to AI substitution. Its advantages — scarce trades, modular prefabrication, repeat relationships and a margin lead over EMCOR — are real but all intact rather than strong, so the franchise is durable through this cycle while the main risk is that the cycle itself digests.
65.0 resilient · 65.0 vulnerable · 80/20 = 65.0 · = 65
Open a moat to read its note.
Construction services contractor; there is no end-user interface to learn.
Mechanical, electrical and modular build know-how is physical execution, not software business logic (M7, the Cameco precedent). Quanta rates this pillar weakened for internal project tooling; under M7 that looks like the one to revisit, not this.
No public-data moat.
Skilled pipefitters, sheet-metal workers and electricians are the binding constraint on data-centre builds, and the decentralised operating companies retain local foremen and estimators well. Rivals such as EMCOR hire from the same craft pool and no captive training pipeline or headcount figure is on file, so it rates intact on M8 — unlike Quanta, which is strong on its captive lineman college.
Mechanical, electrical and factory-built modular scopes let it bid integrated building-systems packages, and Electrical (+81% in Q2) is being cross-sold into mechanical customers. No attach or retention figure is on file, so it rates intact on M9, as Quanta's integrated EPC scope does.
Estimating and job-cost history is internal tooling, not a data asset with a moat. Quanta rates the same class of bid data weakened; under M6 a pillar that never applied is na, so this follows the rule rather than that rating.
State contractor licences, safety records and customer prequalification bar small entrants, but EMCOR and other national contractors clear the same hurdles, so it rates intact on M4, as Quanta's utility prequalification does.
Services contractor with no network effect.
Repeat campus work for hyperscale and advanced-manufacturing builders, plus a service and retrofit business on installed systems, makes it a first-call contractor. Jobs are still bid project by project and EMCOR holds the same relationships, so it rates intact, as Quanta's MSAs do.
Not a system of record.
Modular capacity heading past 4 million square feet and Q2 operating margin of ~17.1% against EMCOR's 10.6% show a real execution and mix advantage, but EMCOR is the larger contractor ($5.15B quarterly revenue vs $3.27B), so the lead is not a unit-cost lead from scale. Intact, as for Quanta.
B2B contractor chosen on capacity, schedule and execution, which the other pillars rate; the name carries reputation, not a price premium it could hold on brand alone (M10).
Growth Analysis
Growth Drivers
Key Risk
Technology work was about 58% of first-half 2026 revenue and much of the backlog sits with a small number of hyperscaler campuses; if two or more of those programmes are paused or re-phased in 2027–28, book-to-bill drops below 1.0, same-store growth falls to single digits, and the ~25–26% gross margin earned in a capacity-tight market reverts toward the low 20s.
Score Derivation
81.3 base − 4.0 trajectory + 4 margin − 5 risk = 76
Base 81.3 (14–20% CAGR, midpoint 17%) − 4 trajectory (Mechanical +47% in Q1 → +40% in Q2; Electrical +88% → +81%; both still very fast but both decelerating) + 4 margin expanding (Q2 operating margin ~17.1% vs ~13.8%; gross margin 25.9% vs 23.5%) − 5 moderate risk (data-centre capex digestion, the same grade as PWR, VRT and ETN) = 76
Price Scenarios (12–24 Months)
Price vs Targets
Dashed lines mark the 12–24 month bear, base and bull targets. Daily closes.
Valuation Multiples
| Trailing P/E (GAAP) | ~42× |
| Forward P/E (FY26) | ~35× |
| Forward P/E (FY27 consensus) | ~28× |
| PEG Ratio | ~1.2× |
| Price / Sales (FY26) | ~4.6× |
At ~28× 2027 consensus FIX trades a little below Quanta (~32× forward) despite a higher operating margin, because its growth is more concentrated in data-centre construction and is decelerating from a 50%+ peak. The ~42× trailing to ~28× two-year-forward gap shows an earnings ramp still under way; the PEG of ~1.2× is fair for growth that holds above 20% into 2027, but leaves the stock exposed if hyperscaler capex digests.
Approximate figures as of October 9, 2026 (price ~$1,715).
Where We Are vs Targets
Loading live price…
Hyperscaler capex digests in 2027, technology orders slow, book-to-bill slips below 1.0 and the multiple compresses toward an E&C norm while 2027 EPS still grows modestly.
- ~22× a reduced 2027 EPS of ~$57 (≈$1,250), a de-rating from ~28× forward, in line with the ~20–24× bear multiples used for Vertiv, Eaton and Quanta
- Two or more hyperscaler campuses are paused or re-phased and backlog falls from the $14.06B record
- Gross margin reverts from ~25.9% toward the low 20s as contractors chase fewer large jobs and the new modular capacity runs below utilisation
2026 EPS lands near $49–50, 2027 grows roughly 20–25% on backlog conversion toward ~$61, and the stock holds a multiple near Quanta's on similar growth.
- ~31× 2027 EPS of ~$61 (≈$1,900), close to Quanta's ~33× base on the same year, ~9% below the ~$2,083 Street mean
- 2026 same-store growth lands in the guided mid-to-high 30s and 2027 revenue reaches ~$15.6B
- Modular capacity reaches ~5M square feet by late summer 2027 and holds gross margin in the mid-20s
AI data-centre and advanced-manufacturing construction extends through 2029, 2028 EPS compounds above $70 and the multiple holds the mid-30s.
- ~36× 2028 EPS of ~$72 (≈$2,600), about 25% above the $2,073.99 52-week high
- Backlog keeps growing past $17B as chip-fab, pharma and power-plant mechanical work adds to data centres
- Electrical keeps compounding through acquisitions like Hunt Electric, and operating margin holds near 17%